Composite valve graft implantation described first in

Transcription

1 Aortic Root Replacement With Composite Valve Graft Davide Pacini, MD, Federico Ranocchi, MD, Emanuela Angeli, MD, Fabrizio Settepani, MD, Marco Pagliaro, MD, Sofia Martin-Suarez, MD, Roberto Di Bartolomeo, MD, and Angelo Pierangeli, MD Department of Cardiac Surgery, University of Bologna, Bologna, Italy Background. Composite valve graft replacement is currently the treatment of choice for a wide variety of lesions of the aortic root and the ascending aorta. In this study we report our experience with aortic root replacement using a composite graft. Methods. Between October 1978 and May 2001, 274 patients (79.6% male and 20.4% female) with a mean age of 53.5 years underwent composite graft replacement of the aortic root. One hundred sixty-one patients (70.8%) had annuloaortic ectasia and 46 (16.8%) aortic dissection. The classic Bentall technique was used in 94 patients (34.3%), the button technique in 172 patients (62.8%), and the Cabrol technique in 8 patients (2.9%). Results. The early mortality rate was 6.9% (19 of 274 patients). Cardiopulmonary bypass time longer than 180 minutes and associated coronary artery bypass grafting were found to be independent risk factors of early mortality. The actuarial survival rate was 77.7% at 5 years and 63% at 10 years. The independent risk factors for late mortality were coronary artery disease, chronic renal failure, and postoperative dialysis. The actuarial freedom from reoperation on the remaining aorta was higher among patients without Marfan syndrome (94.6% versus 79.6% at 10 years, p 0.008). Conclusions. Composite valve graft replacement can be performed with low hospital mortality and morbidity. The button technique offers some advantages and should be used whenever possible. In case of acute aortic dissection root replacement is usually not necessary. Marfan patients should be treated with early root replacement before dissection occurs. (Ann Thorac Surg 2003;76:90 8) 2003 by The Society of Thoracic Surgeons Composite valve graft implantation described first in 1968 by Bentall and De Bono [1] is a welldocumented technique of aortic root replacement used for a large spectrum of pathologic conditions involving the aortic valve and the ascending aorta [2 4]. In the present study we have evaluated the results of our 23-year experience with aortic root replacement (ARR) using a composite valve graft in 274 patients. Material and Methods Patients From October 1978 to May 2001, 274 patients underwent ARR using composite valve graft. Two hundred eighteen patients (79.6%) were male and 56 were female (20.4%). The mean age ( one standard deviation) was years (range, 13 to 80). Thirty-five patients (12.8%) had Marfan syndrome, 2 had Behçet s disease, and 1 had Turner syndrome. Twenty patients (7.3%) were in New York Heart Association (NYHA) functional class I, 84 (30.7%) in functional class II, 113 (41.2%) in functional class III, and 57 (20.8%) in functional class IV. Accepted for publication Feb 4, Address reprint requests to Dr Pacini, c/o Unità Operativa di Cardiochirurgia, Università di Bologna, Policlinico S. Orsola, Via Massarenti, 9, Bologna, Italy; dpacini@hotmail.com. The most common indication for operation was annuloaortic ectasia (161 patients, 58.8%). Thirty-nine patients (14.2%) had previously undergone surgical intervention on the aortic valve or ascending aorta or both. They required reoperation because of progressive dilatation of the Valsalva sinuses in 34, prosthetic aortic valve endocarditis in 4, and acute aortic dissection in 1. The patients profiles are reported in Table 1. Operative Techniques A standard median sternotomy was performed. Cardiopulmonary bypass (CPB) was instituted by cannulation of the ascending aorta, aortic arch, or femoral artery (depending on the extension of the aneurysm and the presence of dissection) and the right atrium or the superior and inferior vena cavae. Myocardial protection was obtained by antegrade administration of cold hyperkalemic crystalloid cardioplegia and topical cooling with 4 C saline solution. For the first 94 patients (34.3%) the classic Bentall operation [1] with inclusion and wrapping technique was used. In 1994 the Bentall procedure was abandoned in favor of the button technique [2 4]. Since then it has been used in 172 patients (62.8%). The coronary reimplantation suture lines were rarely reinforced externally with a Teflon strip. The Cabrol technique [5, 6]was used in 8 patients (2.9%). This method of coronary reimplan by The Society of Thoracic Surgeons /03/$30.00 Published by Elsevier Inc PII S (03)

2 Ann Thorac Surg PACINI ET AL 2003;76:90 8 COMPOSITE VALVE GRAFT REPLACEMENT Table 1. Patient Characteristics Characteristic No. (%) Number of patients 274 Sex Male 218 (79.6) Female 56 (20.4) Age (years) Mean SD Range NYHA class I 20 (7.3) II 84 (30.7) III 113 (41.2) IV 57 (20.8) Marfan syndrome 35 (12.8) Behçet disease 2 (0.7) Indications for operation Primary operation 235 (85.8) Anuloaortic ectasia 161 (68.5) Aortic dissection 46 (19.6) Acute 18 Chronic 28 Poststenotic dilatation 25 (10.6) Endocarditis 3 (1.3) Reoperation 39 (14.2) Aortic valve prosthesis endocarditis 4 (10.3) Valsalva sinus aneurysm after AVR or AAR 34 (87.1) Acute aortic dissection 1 (2.6) AAR ascending aorta replacement or repair; AVR aortic valve replacement; NYHA New York Heart Association. tation was utilized only in case of extreme aortic dilatation or reoperation because of difficult mobilization and approximation of coronary arteries to the aortic graft. In case of acute aortic dissection a hemiarch replacement was usually performed using the open technique. Nevertheless 3 patients required a total arch replacement. In 4 patients the dissection was limitated to the ascending aorta and a closed distal anastomosis was performed. The continuity between the separated layers of the aorta was restored using gelatin-resorcineformaldehyde glue (GRF) and the distal anastomosis was furtherly reinforced with an inner and outer felt strip of Teflon. Concomitant procedures included coronary artery bypass grafting in 23 patients (8.4%), mitral valve replacement in 5 (1.8%), extra-anatomic aorto-aortic bypass in 2, and atrial septal defect repair in 1. Thirty-one patients (11.3%) had associated aortic arch replacement. Cerebral protection was obtained with deep hypothermia with circulatory arrest (DHCA) in 15 patients, DHCA and retrograde cerebral perfusion in 1, and antegrade selective cerebral perfusion with moderate systemic hypothermia in 26 [7]. Mean duration of cardiopulmonary bypass (CPB) was minutes (range, 92 to 425), and mean aortic cross-clamp time was minutes (range, 55 to 305). Table 2. All Variables Analyzed by Univariate Analysis With Respect to Early and Late Mortality Sex Age (13 40; 41 60; 61 70; years) New York Heart Association class (I, II, III, IV) Marfan syndrome Annuloaortic ectasia Aortic dissection (acute, chronic) Endocarditis Associated coronary artery disease Chronic renal failure Reoperation Cardiopulmonary bypass time ( 180, 181 minutes) Clamping time ( 120, 121 minutes) Emergency operation Aortic arch replacement Coronary artery bypass grafting Postoperative dialysis Cardiac complications Postoperative pulmonary insufficiency Postoperative sepsis Postoperative bleeding A Björk-Shiley composite graft prosthesis (Shiley Inc., Irvine, CA) was used in 80 patients (29.2%); a Sorin composite graft (Sorin Biomedica S.P.A., Saluggia, Italy) in 56 (20.4%); a St. Jude composite graft (St. Jude Medical Inc., St. Paul, MN) in 35 (12.8%); a Carbomedics composite graft (Carbomedics Inc., Austin, TX) in 70 (25.6%); and an ATS (ATS Medical Inc., Minneapolis, MN) in 33 (12%). Follow-Up Of all hospital survivors, 239 (93.7%) were available for follow-up in intervals ranging from 3 months to 265 months (mean, 62.7) with a total of 1,431 patients-years. Follow-up information was obtained by our direct examination or by correspondence with the patient. The date of last inquiry was between May and October Postoperative complications were analyzed according to the Guidelines for reporting morbidity and mortality after cardiac valvular operations [8]. Statistical Analysis Statistical analysis was performed with SPSS 8.0 Statistical software (SPSS, Chicago, IL). Continuous variables were expressed as the mean SD and were compared with unpaired two-tailed t test. Categorical variables were analyzed with a 2 test or Fisher s exact test where appropriate. All variables that achieved p less than 0.2 in the univariate analysis were included in a multivariate model and examined by stepwise logistic regression for early mortality, and Cox multivariate analysis for late mortality. All variables analyzed are shown in Table 2. Survival and event-free data were analyzed with Kaplan- Meier actuarial techniques for estimation of survival probabilities and compared with log-rank tests. 91 CARDIOVASCULAR

3 92 PACINI ET AL Ann Thorac Surg COMPOSITE VALVE GRAFT REPLACEMENT 2003;76:90 8 Table 3. Univariate and Multivariate Analysis for Early Mortality Variables Patients Death Univariate No. % No. % p Value Odds Ratio Multivariate 95% Confidence Interval p Value Endocarditis No Yes Associated CAD No Yes Chronic renal failure No Yes CPB time (minutes) Clamping time (minutes) Emergency operation No Yes CABG No Yes Postoperative dialysis No Yes Cardiac complication No Yes Pulmonary insufficiency No Yes Sepsis No Yes CABG coronary artery bypass grafting; CAD coronary artery disease; CPB cardiopulmonary bypass. Results Early Mortality The overall early mortality rate (defined as death within 30 days or during initial hospitalization) was 6.9% (19 of 274 patients). Cause of death was operative myocardial infarction in 5 patients, cardiac arrest in 4, uncontrollable bleeding in 3, multiple organ failure in 2, myocardial failure with impossible weaning from CPB in 2, severe neurologic damage in 1, respiratory insufficiency in 1, and pulmonary thromboembolism in 1. In the univariate analysis (Table 3), coronary artery disease (p 0.009), CPB time (p 0.001), aortic cross-clamp time (p 0.025), associated coronary artery bypass graft surgery ([CABG] p 0.013), cardiac complications (p 0.003), postoperative dialysis (p 0.002), and sepsis (p 0.014) were risk factors for early death. Multivariate analysis indicated CPB time longer than 180 minutes (p 0.001; odds ratio [OR] 12.5) and CABG (p 0.025; OR 4.6) as independent risk factors for early mortality (Table 3). Early Morbidity Cardiac complications occurred in 33 patients (12%) and were associated with an increased risk of early death on univariate analysis. The patients operated on using the Cabrol technique had an high incidence of these complications (3 of 8 [37.5%] versus 30 of 266 [11.3%]). All 3 patients sustained myocardial infarction and 2 of them died. Thirteen patients had persistent or recurrent atrial fibrillation, 8 from complete heart block requiring pacemaker implantation, 3 from ventricular tachycardia/ fibrillation; 1 from myocardial infarction, 1 from endocarditis, 1 from cardiac tamponade, and 1 from left ventricular failure. Fifteen patients (5.5%) sustained respiratory insufficiency requiring prolonged mechanical ventilation (more

4 Ann Thorac Surg PACINI ET AL 2003;76:90 8 COMPOSITE VALVE GRAFT REPLACEMENT Table 4. Causes of Late Deaths Cause Number Congestive heart failure 13 Stroke 6 Prosthetic endocarditis 4 Myocardial infarction 5 Sudden death 11 Rupture of thoracic aorta 4 Cardiac arrest a 1 Suicide 1 Unknown 12 Total 57 a The patient died during operation for abdominal aortic aneurysms. 93 CARDIOVASCULAR than 48 hours). Sepsis occurred in 15 patients (5.5%) and was associated with an increased risk of early mortality (p 0.014). Renal insufficiency requiring dialysis observed in 9 patients (3.3%) was associated with a higher mortality rate (44.4% compared with 5,7%; p 0.002). Nine patients (3.3%) required rethoracotomy for bleeding: 6 (5.9%) underwent the classic Bentall or Cabrol procedure and 3 (1.7%), the button technique. Permanent neurologic deficits developed in 4 patients (1.5%). Late Mortality There have been 57 late deaths (22.3%). The main cause of death was chronic heart failure. The other causes of late death are listed in Table 4. Overall actuarial survival of the 274 patients is shown in Figure 1. The survival rate was 77.7% at 5 years, 63% at 10 years, and 33.4% at 20 years. The survival rate of the patients with Marfan syndrome was lower than that for the remaining patients (61.9% versus 58.8% and 57.7% versus 29.4% at 10 and 15 years respectively) but the difference was not significant (p 0.785; Fig 2A). Moreover Marfan patients with Fig 1. Actuarial survival rates (including hospital mortality) of the 274 patients. Percent survival SE is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years. Number of patients at risk at yearly intervals for years 0 through 20, respectively, is 274, 228, 184, 149, 127, 106, 92, 73, 65, 57, 53, 40, 34, 26, 18, 16, 15, 13, 9, 2, and 2. Fig 2. (A) Actuarial survival rates of the patients with Marfan syndrome (dashed lines) and without Marfan syndrome (solid lines); the difference between the two groups was not significant (p 0.785). Percent survival SE for patients with Marfan syndrome is at 1 year, at 5 years, at 10 years, at 15 years, and 0 at 20 years; number of patients at risk yearly for years 0 through 16, respectively, is 35, 30, 26, 22, 17, 13, 12, 9, 9, 8, 6, 3, 3, 2, 1, 1, and 1. Percent survival for no Marfan syndrome is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years; number of patients at risk yearly for years 0 through 20, respectively, is 239, 198, 156, 126, 110, 92, 79, 64, 56, 49, 47, 37, 30, 24, 17, 15, 14, 14, 9, 2, and 2. (B) Comparison of actuarial survival of patients with aortic dissection (dashed lines) and without aortic dissection (solid lines; p 0.106). Percent survival with aortic dissection is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years; number of patients at risk yearly for years 0 through 20, respectively, is 46, 40, 32, 24, 21, 16, 15, 14, 12, 12, 10, 6, 5, 4, 3, 3, 2, 2, 2, 1, and 1. Percent survival for all other patients is at 1 year, at 5 years, at 15 years, at 15 years, and at 20 years; number of patients at risk yearly for years 0 through 20, respectively, is 288, 188, 150, 124, 106, 88, 76, 59, 52, 45, 43, 33, 28, 22, 15, 13, 12, 12, 7, 1, and 1. dissection demonstrated a 10-year survival of only 42.2% 13.4% whereas no-dissection Marfan patients demonstrated a long-term survival of 64% 26.3%. Patients operated on for aortic dissection had a lower

5 94 PACINI ET AL Ann Thorac Surg COMPOSITE VALVE GRAFT REPLACEMENT 2003;76:90 8 Table 5. Univariate and Multivariate Analysis of Late Mortality Variables Univariate p Value Odds Ratio Multivariate 95% Confidence Interval p Value Preoperative NYHA III/IV Associated CAD Aortic dissection Endocarditis Chronic renal failure Postoperative dialysis NYHA New York Heart Associa- CAD coronary artery disease; tion. long-term survival rate compared with the remaining patients (65% and 58.7% versus 53.6% and 38.1% at 10 and 15 years respectively; Fig 2B). Univariate analysis (Table 5) showed a significant association between late death and NYHA III-IV (p 0.004) associated coronary artery disease (p 0.05), endocarditis (p 0.031), chronic renal insufficiency (p 0.032), and postoperative dialysis (p 0.028). In the Cox multivariate analysis, associated CAD (p 0.028; OR 2.3), chronic renal failure (p 0.012, OR 4.0), and postoperative dialysis (p 0.039; OR 2.9) were independent risk factors for late mortality (Table 5). Fig 3. (A) Actuarial freedom from thromboembolism. Percent of patients free of thromboembolism SE is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years. Number of patients at risk yearly for years 0 through 20, respectively, is 274, 228, 182, 147, 126, 103, 89, 73, 65, 57, 52, 39, 32, 25, 17, 15, 14, 13, 8, 1, and 1. (B) Actuarial freedom from anticoagulant-related hemorrhage. Percent of patients free of hemorrhage is 100 at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years. Number of patients at risk yearly for years 0 through 20, respectively, is 274, 228, 183, 148, 127, 105, 91, 72, 63, 55, 51, 40, 33, 26, 18, 15, 14, 13, 8, 2, and 2. Late Morbidity Thromboembolic events (TE) occurred in 9 patients (3.3%) and all of them had a stroke. Two patients died. The linearized rate of TE was 0.63/100 patient-years. Figure 3A shows the actuarial freedom from TE. At 15 years the actuarial freedom from TE was 90.9% 3.1%. Thirteen patients (4.7%) had anticoagulant-related bleeding events necessitating hospital admission or blood transfusion or resulting in death. Five patients had cerebral hemorrhage and 4 of them died. Eight patients had gastrointestinal bleeding or retroperitoneal hematoma or both. The linearized risk of anticoagulant-related hemorrhage was 0.91/100 patient-years. Estimates for freedom from bleeding complications are shown in Figure 3B. In 5 patients (1.8%) prosthetic valve endocarditis developed (1 early and 4 late). Two patients underwent reoperation (1 died) and 3 were treated with medical therapy (1 survivor). The linearized risk per 100 patientyears of prosthetic valve endocarditis (early and late) was The actuarial freedom from endocarditis is shown in Figure 4A. Four patients underwent reoperation for pseudoaneurysm of the coronary ostial suture line. In 2 of them the classic Bentall operation was used (2.1%) and in the other 2 (1 with Behçet s disease) the button technique was used (1.2%). The linearized rate of reoperation for pseudoaneurysm was 0.28/100 patient-years. Freedom from reoperation due to pseudoaneurysm for the classic Bentall and for the button technique was, respectively, 98.8% and 98.8% at 5 years, 98.8% and 95.5% at 10 years (Fig 4B). The difference between the two groups was not significant (p 0.776). Thromboemboic events, anticoagulant-related hemorrhage, prosthetic valve endocarditis, and reoperations for pseudoaneurysms were reviewed to evaluate overall valve graft-related morbidity. The actuarial estimate of percentage of patients free of any valve graft-related complications is shown in Figure 5. Eleven patients have required one or more subsequent

6 Ann Thorac Surg PACINI ET AL 2003;76:90 8 COMPOSITE VALVE GRAFT REPLACEMENT 95 CARDIOVASCULAR Fig 5. Actuarial freedom from valve graft related complications. Percent of patients free of complications SE is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years. Number of patients at risk yearly for years 0 through 20, respectively, is 274, 226, 182, 147, 126, 103, 89, 72, 63, 54, 49, 39, 32, 25, 17, 14, 13, 12, 7, 1, and 1. Comment Since its introduction in 1968 by Bentall and De Bono [1] the aortic valve and ascending aorta replacement with composite graft has led to a significant prolongation of life expectancy for patients affected by a variety of pathologic conditions involving the ascending aorta and aortic valve such as annuloaortic ectasia, cystic medial necrosis Fig 4. (A) Actuarial freedom from prosthetic endocarditis. Percent of patients free of prosthetic endocarditis SE is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years. Number of patients at risk yearly for years 0 through 20, respectively, is 274, 225, 181, 146, 125, 102, 88, 71, 62, 53, 42, 39, 32, 25, 17, 14, 13, 12, 7, 1, and 1. (B) Actuarial freedom from reoperation for pseudoaneurysms according to the operative technique: classic Bentall technique (dashed lines) and button technique (solid lines; p 0.776). Percent of Bentall patients free of pseudoaneurysms is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years; number of patients at risk yearly for years 0 through 20, respectively, is 94, 83, 77, 75, 73, 70, 68, 68, 63, 56, 53, 40, 33, 26, 18, 16, 15, 14, 9, 2, and 2. Percent of button technique patients free of pseudoaneurysms is 100 at 1 year, at 5 years, and at 10 years; number of patients at risk at yearly intervals for years 0 through 8, respectively, is 172, 140, 101, 70, 51, 31, 21, 4, and 1. interventions for aneurysm or dissection of the remaining aorta; 4 were Marfan patients and all of them had aortic dissection. The rate of freedom from aortic reoperation of the patients with Marfan syndrome was lower than that for the remaining patients at 10 years (79.6% 13.6% versus 94.6% 2.6%) and the difference was significant (p 0.008; Fig 6). Fig 6. Actuarial freedom from reoperation on the thoracic or abdominal aorta or both in patients with Marfan syndrome (dashed lines) and without Marfan syndrome (solid lines); the difference between the two groups was statistically significant (p 0.008). Percent of Marfan patients free of reoperation SE is 100 at 1 year, at 5 years, at 10 years, at 15 years, and 0 at 20 years; number of patients at risk yearly for years 0 through 16, respectively, is 35, 30, 26, 22, 17, 13, 12, 9, 9, 7, 6, 3, 3, 2, 1, 1, and 1. Percent of non-marfan patients free of reoperation is at 1 year, at 5 years, at 10 years, at 15 years, and at 20 years; number of patients at risk yearly for years 0 through 20, respectively, is 239, 198, 157, 126, 110, 92, 79, 64, 56, 49, 46, 37, 30, 24, 17, 15, 14, 14, 8, 2, and 2.

7 96 PACINI ET AL Ann Thorac Surg COMPOSITE VALVE GRAFT REPLACEMENT 2003;76:90 8 with or without Marfan syndrome, and type A aortic dissection [3, 4, 9 17]. Our retrospective analysis confirms, in agreement with other recent reports [14 17], that this surgical procedure presents a low operative risk. We found CPB time longer than 180 minutes and associated CABG to be independent risk factors of early mortality. In 1994 we abandoned the original Bentall operation with the inclusion technique and introduced the button technique [2 4] with several features that may reduce the incidence of early and late complications. Hemostasis may be improved by avoiding aortic wall wrapping. Since this modification repeat thoracotomy for bleeding has been reduced from 5.9% to 1.7%. Other factors such as the use of preclotting woven aortic graft, improved surgeon experience, and more accurate use of eparine and protamine may have contributed to the reduction of intraoperative bleeding. The button technique without complete aortic wall wrapping may prevent late pseudoaneurysm formation [2 4] secondary to dehiscence of the suture line of the aortic annulus, distal graft anastomosis, or mainly at the coronary ostial anastomosis particularly at the left coronary ostium in patients with Marfan syndrome or with aortic dissection [12, 18]. Kouchoucos and associates [3] have suggested that blood accumulation within the wrapped perigraft space results in increased tension on the anastomosis when the inclusion and wrap technique is used. We also believe that the circumferential resection of the distal part of the ascending aorta reinforced with two Teflon strips can reduce the stress along the suture line between composite prosthesis and aorta. We rarely performed coronary reimplantation according to the Cabrol technique [5, 6]; it was used for patients who had undergone reoperation or for cases of extreme aortic dilatation because of difficult mobilization and approximation of the coronary arteries to the aortic graft. One of the greatest technical difficulties with this technique is the sizing and orienting of the graft between the right and left main coronary arteries to prevent kinking and subsequent myocardial ischemia or infarction. In our experience it was associated with a high early mortality rate (3 of 8, 37.5%) and a high incidence of perioperative myocardial infarction. Therefore since 1994 we have not used the Cabrol technique, and detachment and mobilization of the coronary arteries could be easily performed using the button technique. As cardiovascular manifestations are the most important causes of death among Marfan patients, the survival of these patients depends on the prevention and control of these complications. A recent study [19] showed life expectancy improvement among patients with Marfan syndrome who had undergone surgical repair for aortic aneurysms. Now the median cumulative probability of survival is 61 years whereas 30 years ago, it was 47 years. In our series the survival rate including 30-day mortality of the Marfan patients was slightly lower than that of the remaining patients (at 10 years 58.8% 14.6% versus 61.9% 4.6%) without statistical difference (p 0.785). A factor that influenced the late survival of the Marfan patients was the presence of aortic dissection: at 10 years the survival rate of the Marfan patients with aortic dissection was 42.2% whereas that for the Marfan patients without dissection was 64%. Late mortality was associated with associated CAD, chronic renal failure, and postoperative dialysis. Although some researchers have found long-term survival to be statistically less favorable among patients with aortic dissection at the time of root replacement [16] it was not a predictor of late mortality in our series. That may be due to a low rate of patients with Marfan syndrome in our series (12.8%) compared with that reported in the literature (69.3%) [16]. Crawford [2] underlined in a large series of patients with dissection or aneurysm of the ascending aorta or aortic arch that diseases of the aorta are often part of a more diffuse degenerative process. The same author [20] reported an elevated incidence of operation on the remaining aorta among patients with Marfan syndrome who underwent composite graft or aortic valve replacement. A recent paper [21] confirmed a significant progression of the disease in the remaining aorta in Marfan patients who had previously undergone composite graft replacement. Seventeen of the 48 patients studied by magnetic resonance imaging had a significant increase in diameter of the aorta with a mean rate of dilation of mm per year. Surgical intervention was necessary in 14 of them. In our study the rate of freedom from reoperation on the remaining aorta of the patients with Marfan syndrome was lower than of the other patients at 10 years (79.6% 13.6% versus 94.6% 2.6%) and the difference was significant (p 0.008). Moreover all Marfan patients reoperated on during follow-up had aortic dissection. According to this, all patients who have undergone aortic root replacement should be periodically evaluated by computed tomography scan, magnetic resonance imaging, or transesophageal echocardiography to detect the development of false aneurysms or the progression of the disease in the remaining aorta, particularly in patients with Marfan syndrome or with aortic dissection. Despite refinements in the design of cardiac prostheses and in anticoagulation management, mechanical valve replacement is still associated with a variety of valverelated complications often leading to serious disability or death. In our series the rate of valve-related complications was low. Anticoagulant-related hemorrhage was the most common late complication with a rate of 0.91 events per 100 patient-years, followed by thromboembolisms (0.63/100 patient-years). Endocarditis was a serious complication with a high mortality rate (60%). It is our standard policy to replace the infected composite graft or prosthesis with a cryopreserved homograft root. Conservative treatment failed to eradicate infection in all patients treated except for 1 patient. Appropriate antibiotic prophylaxis remains the main preventative measure. To avoid the disadvantages of prosthetic heart valves the valve-sparing procedure has been introduced [22]. Patients with aortic root aneurysm often have normal or minimally diseased aortic cusps that can be preserved.

8 Ann Thorac Surg PACINI ET AL 2003;76:90 8 COMPOSITE VALVE GRAFT REPLACEMENT Actually the valve-sparing operation has become our treatment of choice for aortic root aneurysm with normal aortic valve and in the past 24 months we have performed 24 procedures. However the current series does not include these patients and it reports only our experience with composite valve graft replacement. Four patients (1.5%) required reoperation for pseudoaneurysm formation at the coronary suture lines: in 2 patients the original Bentall operation was used (2.1%) and in the other 2 1 of them with Behçet s disease the button technique was used (1.2%). All patients underwent successful reoperation. Techniques used for reattachment of coronary arteries did not influence the incidence of reoperation for pseudoaneurysm during follow-up. Because not all patients were evaluated by diagnostic imaging studies such as magnetic resonance, computed tomography, or angiography the real incidence of pseudoaneurysm formation is unknown and may be higher. When a pseudoaneurysm is detected it should be repaired before progressive dilation, adherence to the sternum, or rupture because all these situations, which require urgent or emergent operation, are associated with high operative risk [3]. Hahn and associates [23] reported no early deaths in a limited series of patients who had undergone aortic root reoperation for pseudoaneurysm or endocarditis but no operations were done emergently. In a study of 81 patients who had undergone reoperation on the aortic root or ascending aorta Kouchoukos and colleagues [24] presented an early mortality rate of 12.5% in 16 patients reoperated on for false aneurysm. In the same report reoperation for false aneurysm was a significant predictor of late mortality. Eighteen of the 274 patients (66%) who underwent aortic root replacement had acute type A dissection. This number represents fewer than 10% of all patients operated on for acute type A dissection during the same period. We believe along with Elefteriades [25] that the vast majority of aortic dissections can be treated appropriately with a simple supracoronary hemiarch replacement and the aortic valve can be left alone or the commissures can be resuspended. Long-term survival after root replacement for acute aortic dissection was found to be statistically less favorable [10]. Composite graft replacement should be limitated to cases of frank annuloaortic ectasia, Marfan syndrome, and severe destruction of the proximal aorta. Finally we should mention some limitations of the current investigation. First, this is a retrospective study over a long period of time in which many factors changed and could not be accounted for with the multivariate statistical techniques. Second, owing to incomplete data collection during the earlier years some important variables such as left ventricular function and intraoperative myocardial protection were not included in the analysis. Therefore the influence of these variables on early and late mortality could not be studied. In summary composite valve graft replacement can be performed with low rates of hospital mortality and morbidity. The button technique offers some advantages and should be used whenever possible. In case of acute aortic dissection root replacement is usually not necessary. Patients with Marfan syndrome should undergo early root replacement before aortic dissection occurs. Valverelated complications have a low incidence but often lead to disability or death. A careful follow-up is extremely important for evaluating the prosthetic aortic segment, the proximal and distal anastomosis, the morphology, and the diameter of the reimplantated coronary arteries and the remaining segments of the aorta. References 1. Bentall HH, De Bono A. A technique for complete replacement of the ascending aorta. Thorax 1968;23: Crawford ES, Svensson LG, Coselli JS, Safi HJ, Hess KR. Surgical treatment of aneurysm and/or dissection of the ascending aorta, transverse aortic arch, and ascending aorta and transverse aortic arch. Factors influencing survival in 717 patients. J Thorac Cardiovasc Surg 1989;98: Kouchoukos NT, Wareing TH, Murphy SF, Perillo JB. Sixteen-year experience with aortic root replacement. Results of 172 operations. Ann Surg 1991;214: Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Composite valve graft replacement of the proximal aorta: comparison of techniques in 348 patients. Ann Thorac Surg 1992;54: Cabrol C, Pavie A, Gandjbakhch I, et al. Complete replacement of the ascending aorta with reimplantation of the coronary arteries. New surgical approach. J Thorac Cardiovasc Surg 1981;81: Cabrol C, Pavie A, Mesnildrey P, et al. Long-term results with total replacement of the ascending aorta and reimplantation of the coronary arteries. J Thorac Cardiovasc Surg 1986;91: Di Bartolomeo R, Pacini D, Di Eusanio M, Pierangeli A. Antegrade selective cerebral perfusion during operations on the thoracic aorta: our experience. Ann Thorac Surg 2000;70: Edmunds LH, Clark RE, Cohn LH, Grunkemeier GL, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. J Thorac Cardiovasc Surg 1996;112: Lewis CTP, Cooley DA, Murphy MC, Talledo O, Vega D. Surgical repair of aortic root anurysm in 280 patients. Ann Thorac Surg 1992;53: Lytle BW, Mahfood SS, Cosgrove DM, Loop FD. Replacement of the ascending aorta: early and late results. J Thorac Cardiovasc Surg 1990;99: Marsalese DL, Moodie DS, Vacane M, et al. Marfan s syndrome: natural history and long-term follow-up of cardiovascular involvement. J Am Coll Cardiol 1989;14: Svensson LG, Crawford ES, Coselli JS, Safi HJ, Hess KR. Impact of cardiovascular operation on survival in the Marfan patient. Circulation 1989;80(Suppl 1): Gott VL, Pyeritz RE, Cameron DE, Greene PS, MuKusick VA. Composite graft repair of Marfan aneurysm of the ascending aorta: results of 100 patients. Ann Thorac Surg 1991;52: Mingke D, Dresler C, Pethig K, Heinemann M, Borst HG. Surgical treatment of Marfan patients with aneurysms and dissection of the proximal aorta. J Cardiovasc Surg 1998;39: Niederhauser U, Rudiger H, Vogt P, Kunzli A, Zund G, Turina M. Composite graft replacement of the aortic root in acute dissection. Eur J Cardiothorac Surg 1998;13: Gott VL, Gillinov AM, Pyeritz RE, et al. Aortic root replacement. Risk factor analysis of a seventeen-year experience with 270 patients. J Thorac Cardiovasc Surg 1995;109: Mingke D, Dresler C, Stone CD, Borst HG. Composite graft 97 CARDIOVASCULAR

9 98 PACINI ET AL Ann Thorac Surg COMPOSITE VALVE GRAFT REPLACEMENT 2003;76:90 8 replacement of the aortic root in 335 patients with aneurysm or dissection. Thorac Cardiovasc Surg 1998;46: Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Dissection of the aorta and dissecting aortic aneurysms: improving early and long term results. Circulation 1990; 82(Suppl 4): Finkbohner R, Johnston D, Crawford ES, Coselli J, Milewicz D. Marfan syndrome: longterm survival and complications after aortic aneurysm repair. Circulation 1995;91: Crawford ES. Marfan s syndrome. Broad spectral surgical treatment of cardiovascular manifestations. Ann Surg 1983; 198: Kawamoto S, Bluemke DA, Traill TA, Zerhouni EA. Thoracoabdominal aorta in Marfan syndrome: MR imaging findings of progression of vasculopathy after surgical repair. Radiology 1997;203: David TE, Feindel M. An aortic valve-sparing operation for patients with aortic incompetence and aneurysm of the ascending aorta. J Thorac Cardiovasc Surg 1992;103: Hahn C, Tam SK, Vlahakes GJ, Hilgenberg AD, Akins CW, Buckley MJ. Repeat aortic root replacement. Ann Thorac Surg 1998;66: Dougenis D, Daily BB, Kouchoukos NT. Reoperations on the aortic root and ascending aorta. Ann Thorac Surg 1997;64: Elefteriades JA. What operation for acute type A dissection? J Thorac Cardiovasc Surg 2002;123:201 3.